A dynamic force feedback compensation method for robot compliant teleoperation

Abstract

The invention discloses a dynamic force feedback compensation method for compliant remote operation of a robot, comprising establishing a compensation mapping relationship between input work and output work, calculating the output work of the robot and the input work of an operator in a unit time and scalarizing the input work of the operator in Cartesian space coordinates, and solving the compensation force of the remote operation device at each moment by using the cosine theorem, the functional compensation mapping relationship and iterative calculation, thereby realizing dynamic compensation of the feedback force of the remote operation device for the impact that may be suffered during the movement of the robot arm, and avoiding the impact and vibration in real time; enabling the operator to clearly and explicitly perceive the impact; in the process of iteratively solving the compensation force, under the premise of insufficient linear equations, using the directional relationship between the feedback force and the speed of the operator operating the remote operation device to provide conditions for solving using the cosine theorem, which can effectively improve the smoothness of the trajectory of the remote operation robot and reduce the impact and vibration suffered by the robot during the movement.

Description

Technical Field

[0001] The invention relates to the technical field of teleoperation robot control, in particular to a dynamic force feedback compensation method applied to robot compliant teleoperation. Background Art

[0002] Teleoperation is one of the earliest research directions and application fields of robotics. This operation technology can complete complex tasks by remotely controlling robots in complex environments. Teleoperation technology overcomes the spatial barriers of the operator's operation of the robot, making robots widely used in more dangerous scenarios (such as the nuclear industry, chemical industry, etc.) and scenarios where operators cannot go in person (such as minimally invasive surgery, telemedicine, etc.).

[0003] Among the current teleoperated robot control methods, the position-based direct control method is the most commonly used method. The position information is sent in real time by the artificial control end, and the robot execution end moves according to the position information. In the actual complex application process, the motion information output by the control end is usually not trajectory planned, and is not flexible and continuous. After position scaling and power amplification, these position information will bring severe rigid and flexible impacts to the robot movement, which brings great hidden dangers to the stability and safety of the teleoperation system.

[0004] The jitter problem of teleoperated robots is increasingly becoming an issue that needs to be urgently solved in the application of this technology in more fields, but there is currently little related research work; therefore, further improving the direct control method of the robot of the teleoperated system to avoid the jitter problem of the robot during movement as much as possible has great research significance and necessity for the application of teleoperated robot technology and the expansion of application scenarios. Summary of the invention

[0005] The present invention provides a dynamic force feedback compensation method for robot compliant teleoperation. The operator senses the compensated feedback force through the teleoperation device, thereby avoiding impact and vibration in real time, which can effectively improve the smoothness of the trajectory of the teleoperated robot and reduce the impact and vibration received by the robot during movement.

[0006] To achieve the above object, the present invention provides the following technical solution: a dynamic force feedback compensation method applied to robot compliant teleoperation, comprising the following steps:

[0007] S1. Obtain the robot output work ΔW per unit time s and the operator input work ΔW m , where ΔW m Compensation adjustment is ΔW m-comp ;

[0008] S2, establish ΔW s With ΔW m-comp The mapping relationship between them is:

[0009]

[0010] in, is the rigid impact linear amplification function, is the linear amplification function of the flexible impact, k R-joint is the joint weighting coefficient of rigid impact, k F-joint is the joint weighting coefficient of flexible impact;

[0011] S3, approximate calculation of the robot output work ΔW per unit time through differential thinking s and the operator input work ΔW m_comp And input work ΔW to the operator m-comp Scalarization in Cartesian space coordinates;

[0012] S4, based on ΔW s With ΔW m-comp The mapping relationship between them is calculated by using the cosine theorem and iterative calculation to obtain the compensation force F of the teleoperation device at each moment. m-comp .

[0013] Preferably, in step S1, based on an arbitrary time interval Δt, the displacement p of the mapped motion point of the teleoperated device is used as the integral variable to calculate the feedback force F on the teleoperated device. m Integrate to get the operator input work per unit time ΔW m :

[0014] Taking the rotation angle q on the robot joint as the integral variable, the rotation angle τ on the robot joint is integrated to obtain the robot output work ΔW per unit time: s The expression is:

[0015] Among them, m represents the teleoperated device controlled by the operator, and s represents the robot.

[0016] Preferably, in step S1, ΔW m-comp Greater than ΔW m .

[0017] Preferably, in step S2, the robot has a degree of freedom of n, a rigid impact proportional coefficient a and a flexible impact proportional coefficient b;

[0018] in, It is an n×1 vector formed by the joint angular acceleration of each joint of the robot;

[0019] in, It is an n×1 vector formed by the joint angular acceleration derivatives of each joint of the robot.

[0020] Preferably, in step S3, the robot output work ΔW s and the operator input work ΔW m-comp Respectively expressed as vector dot product form:

[0021]

[0022]

[0023] Among them, comp represents the data after compensation processing.

[0024] Preferably, in step S3, the operator is input with work ΔW m-comp Scalarization is performed in Cartesian space coordinates, specifically:

[0025] ΔW m-comp The dot product definition converts the vector dot product form into the form of multiplying the vector magnitude and the vector angle:

[0026] Among them, θ m-comp Represents the vector [F m_comp (t+Δt)+F m_comp (t)] and the angle formed by the vector [p(t+Δt)-p(t)], where

[0027] Decomposed in Cartesian space:

[0028]

[0029]

[0030] p(t)=[x p (t),y p (t),z p (t)] T ;

[0031] p(t+Δt)=[x p (t+Δt),y p (t+Δt),z p (t+Δt)] T .

[0032] Preferably, in step S4, based on ΔW s With ΔW m-comp The mapping relationship between |F m_comp(t+Δt)+F m_comp (t)|Separation:

[0033] a. Set ΔW s Scalarization: Among them, θ s represents the angle between the vector [τ(t+Δt)+τ(t)] and the vector [q(t+Δt)-q(t)], where cosθ m_comp ≈1,

[0034] b. Scale the scalared ΔW s Substitute ΔW s With ΔW m-comp The mapping relationship between them will be m_comp (t+Δt)+F m_comp (t)|Out:

[0035]

[0036] Preferably, in step S4, let α be the vector F m_comp (t) and vector F m_comp (t+Δt) and the velocity vector of the operator operating the teleoperated device and Opposite direction:

[0037]

[0038] Among them, the vector [F m_comp (t+Δt)+F m_comp (t)] is in the vector F m_comp (t), F m_comp On the bisector of the angle formed by (t+Δt), according to the cosine theorem, the vector F m_comp (t) and the vector [F m_comp (t+Δt)+F m_comp (t)]

[0039]

[0040] |F m_comp (t+Δt)| for sorting and separation:

[0041]

[0042] Vector|F m-comp The direction of (t+Δt)| Similarly, through iterative calculation, the compensation force F of the teleoperation device at each moment is obtained m-comp .

[0043] Compared with the prior art, the present invention has the following beneficial effects: the compensated feedback force is sensed by the teleoperation device, thereby avoiding impact and vibration in real time; a compensation mapping between the operator's input work and the robot's output work is established through a functional relationship, so that the operator can clearly and explicitly sense the impact; in the process of iteratively solving the compensation force, under the premise that the linear equation group is insufficient, the directional relationship between the feedback force and the operator's operating speed of the teleoperation device is used to provide conditions for solving the problem using the cosine theorem, which can effectively improve the smoothness of the teleoperation robot's trajectory and reduce the impact and vibration received by the robot during movement. In addition, the mapping relationship established based on the functional relationship also makes it easier for the teleoperation system to perform steady-state analysis using Lyapunov's second method. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0045] In the attached picture:

[0046] Figure 1 It is a schematic diagram of the structure of the compliant teleoperation of the robot of the present invention;

[0047] Figure 2 It is a vector relationship diagram of the compliant teleoperation of the robot of the present invention. DETAILED DESCRIPTION

[0048] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0049] Example: Figure 1-Figure 2 As shown, a dynamic force feedback compensation method for robot compliant teleoperation, wherein for a complete teleoperation robot system, including a robot with n degrees of freedom, an industrial computer for communication based on a real-time operating system of ROS, and a high-precision teleoperation device (such as an omega.7 teleoperation lever), the following steps are included:

[0050] S1. Based on an arbitrary time interval Δt, the displacement p of the mapped motion point of the teleoperated device is used as the integral variable to calculate the feedback force F on the teleoperated device. m Integrate to get the operator input work per unit time ΔW m :

[0051] Taking the rotation angle q on the robot joint as the integral variable, the rotation angle τ on the robot joint is integrated to obtain the robot output work ΔW per unit time: s The expression is:

[0052] Among them, m represents the teleoperated device controlled by the operator, and s represents the robot;

[0053] Based on the selected sufficiently small time interval Δt, the industrial computer communicates with the robot once, including reading the velocity information and acceleration information on the robot joints, and further obtaining the acceleration derivative information on the robot joints through differential operations; ΔW m Compensation adjustment is ΔW m_comp , in order to enhance the feedback effect of the operator's perception of the robot's impact when using the teleoperated device, the compensated ΔW m-comp Should be compared to ΔW m Bigger;

[0054] Set the rigid impact proportional coefficient a and the flexible impact proportional coefficient b. These two coefficients are usually set according to actual operation experience; then construct the rigid impact linear amplification function And the flexible impact linear amplification function In form, it is a vector of n×1, based on which the value of the linear amplification function per unit time is calculated:

[0055] in, It is an n×1 vector formed by the joint angular acceleration of each joint of the robot;

[0056] in, It is an n×1 vector formed by the joint angular acceleration derivatives of each joint of the robot.

[0057] S2, considering that when different joints of the robot are subjected to the same impact, the robot as a whole will show inconsistent vibration effects, so the n×1 weighted coefficient vector of each joint is set, and the rigid impact and flexible impact are expressed as k R-joint , k F-joint ;

[0058] Establish ΔW s With ΔW m-comp The mapping relationship between them is:

[0059]

[0060] in, is the rigid impact linear amplification function, is the linear amplification function of the flexible impact, k R-joint is the joint weighting coefficient of rigid impact, k F-joint is the joint weighting coefficient of flexible impact;

[0061] The compensation mapping relationship established above can convert the robot output work ΔW generated by rigid impact and flexible impact into s Clearly and unambiguously mapped to operator input work ΔW m-comp In general, after the compensation force is separated, the operator can feel the corresponding compensation force on the teleoperation device; in addition, the mapping relationship established based on the functional relationship also makes it easier for the teleoperation system to perform mainstream steady-state analysis.

[0062] S3, approximate calculation of the robot output work ΔW per unit time through differential thinking s and the operator input work ΔW m-comp And input work ΔW to the operator m_comp Scalarization in Cartesian space coordinates;

[0063] a. Take Δt as a sufficiently small value, and for the robot output work ΔW s and the operator input work ΔW m-comp Respectively expressed as vector dot product form:

[0064]

[0065]

[0066] Among them, comp represents the data after compensation processing;

[0067] b. Input work ΔW to the operator m-comp Scalarization is performed in Cartesian space coordinates, specifically:

[0068] ΔW m-comp The dot product definition converts the vector dot product form into the form of multiplying the vector magnitude and the vector angle:

[0069] Among them, θ m-comp Represents the vector [F m_comp (t+Δt)+F m_comp (t)] and the angle formed by the vector [p(t+Δt)-p(t)], where the compensation force F of the teleoperation device m_comp (t) is always related to the speed of movement during the movement The directions are opposite, so when Δt is small enough, the vector [F m_comp (t+Δt)+F m_comp (t)] will be related to the vector [p(t+Δt)-p(t)], the velocity vector at time t The direction is approximately opposite, that is, cosθ m_comp ≈-1,

[0070] Decomposed in Cartesian space:

[0071]

[0072]

[0073] p(t)=[x p (t),y p (t),z p (t)] T ;

[0074] p(t+Δt)=[x p (t+Δt),y p (t+Δt),z p (t+Δt)] T .

[0075] S4, based on ΔW s With ΔW m-comp The mapping relationship between them is used, and the compensation force F at the next moment is calculated by using the cosine theorem and iterative calculation. m_comp (t+Δt) Input work F from the operator m_comp (t+Δt) separation, obtain the compensation force F of the teleoperated device at each moment m-comp :

[0076] a. Based on ΔW s With ΔW m-comp The mapping relationship between |F m_comp (t+Δt)+F m_comp (t)|Separation:

[0077] (1) ΔW s Scalarization: Among them, θ s represents the angle between the vector [τ(t+Δt)+τ(t)] and the vector [q(t+Δt)-q(t)], where the robot’s torque τ(t) is always proportional to the joint velocity during motion. The direction is the same, so when Δt is small enough, the vector [τ(t+Δt)+τ(t)] will be the same as the vector [q(t+Δt)-q(t)], the velocity vector at time t The directions are approximately the same, that is,

[0078] b. Scale the scalared ΔW s Substitute ΔW s With ΔW m-comp The mapping relationship between them will be m_comp (t+Δt)+F m_comp (t)|Out:

[0079]

[0080] b. Let α be the vector F m_comp (t) and vector F m_comp (t+Δt) and the velocity vector of the operator operating the teleoperated device and Opposite direction:

[0081]

[0082] In which, according to the parallelogram law, the vector [F m_comp (t+Δt)+F m_comp (t)] is in the vector F m_comp (t), F m_comp On the bisector of the angle formed by (t+Δt), according to the cosine theorem, the vector F m_comp (t) and the vector [F m_comp (t+Δt)+F m_comp (t)]

[0083]

[0084] |F m_comp (t+Δt)| for sorting and separation:

[0085]

[0086] Vector|F m-comp The direction of (t+Δt)| Same, complete the F m_comp Separation and solution of (t+Δt), through iterative calculation, obtain the compensation force F of the teleoperation equipment at each moment m-comp The industrial computer transmits the compensation force information to the teleoperation device in real time to complete the feedback force compensation process. Under the premise that the linear equations are insufficient, the directional relationship between the feedback force and the operator's teleoperation device speed is used to separate F using the cosine theorem. m_comp (t+Δt) provides the conditions.

[0087] Finally, it should be noted that the above description is only a preferred example of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A dynamic force feedback compensation method for robot compliant teleoperation, characterized in that: The steps include: S1. Obtain the robot output work ΔW per unit time s and the operator input work ΔW m , where ΔW m Compensation adjustment is ΔW m-comp ; Based on an arbitrary time interval Δt, the displacement p of the mapped motion point of the teleoperated device is taken as the integral variable to calculate the feedback force F on the teleoperated device. m Integrate to get the operator input work per unit time ΔW m : Taking the rotation angle q on the robot joint as the integral variable, the rotation angle τ on the robot joint is integrated to obtain the robot output work ΔW per unit time: s The expression is: Among them, m represents the teleoperated device controlled by the operator, and s represents the robot; ΔW m-comp Greater than ΔW m ; S2, establish ΔW s With ΔW m-comp The mapping relationship between them is: in, is the rigid impact linear amplification function, is the linear amplification function of flexible impact, k R-joint is the joint weighting coefficient of rigid impact, k F-joint is the joint weighting coefficient of flexible impact; The robot has n degrees of freedom, a rigid impact proportionality coefficient and b flexible impact proportionality coefficient; in, It is an n×1 vector formed by the joint angular acceleration of each joint of the robot; in, It is an n×1 vector formed by the derivatives of the joint angular acceleration of each joint of the robot; S3, approximate calculation of the robot output work ΔW per unit time through differential thinking s and the operator input work ΔW m-comp And input work ΔW to the operator m-comp Scalarization in Cartesian space coordinates; The robot output work ΔW s and the operator input work ΔW m-comp Respectively expressed as vector dot product form: Among them, comp represents the data after compensation processing; S4, based on ΔW s With ΔW m-comp The mapping relationship between them is calculated by using the cosine theorem and iterative calculation to obtain the compensation force F of the teleoperation device at each moment. m-comp .

2. A dynamic force feedback compensation method for robot compliant teleoperation according to claim 1, characterized in that: In step S3, the operator is asked to input work ΔW m-comp Scalarization is performed in Cartesian space coordinates, specifically: ΔW m-comp The dot product definition converts the vector dot product form into the form of multiplying the vector magnitude and the vector angle: Among them, θ m-comp Represents the vector [F m_comp (t+Δt)+F m_comp (t)] and the angle formed by the vector [p(t+Δt)-p(t)], where cosθ m_comp ≈-1, Decomposed in Cartesian space: p(t)=[x p (t),y p (t),z p (t)] T ; p(t+Δt)=[x p (t+Δt),y p (t+Δt),z p (t+Δt)] T 。 3. A dynamic force feedback compensation method for robot compliant teleoperation according to claim 2, characterized in that: In step S4, based on ΔW s With ΔW m-comp The mapping relationship between |F m_comp (t+Δt)+F m_comp (t)|Separation: a. Set ΔW s Scalarization: Among them, θ s represents the angle between the vector [τ(t+Δt)+τ(t)] and the vector [q(t+Δt)-q(t)], where b. Scale the scalared ΔW s Substitute ΔW s With ΔW m-comp The mapping relationship between them will be m_comp (t+Δt)+F m_comp (t)|Out:

4. The dynamic force feedback compensation method for robot compliant teleoperation according to claim 3, characterized in that: In step S4, let α be the vector F m_comp (t) and vector F m_comp (t+Δt) and the velocity vector of the operator operating the teleoperated device and Opposite direction: Among them, the vector [F m_comp (t+Δt)+F m_comp (t)] is in the vector F m_comp (t), F m_comp On the bisector of the angle formed by (t+Δt), according to the cosine theorem, the vector F m_comp (t) and the vector [F m_comp (t+Δt)+F m_comp (t)] |F m_comp (t+Δt)| for sorting and separation: Vector|F m-comp The direction of (t+Δt)| Similarly, through iterative calculation, the compensation force F of the teleoperation device at each moment is obtained m-comp .

Images